Patients with HIV often develop a painful neuropathy that is exacerbated by treatment with highly active anti-retroviral therapy (HAART). The recent development of animal models, created by application of gp120 to the sciatic nerve along with systemic administration of nucleoside reverse transcriptase inhibitors (NRTIs), have allowed studies of the pathogenesis of HIV/HAART neuropathic pain. We and others have described an increase in tumor necrosis factor (TNFa) in dorsal root ganglia (DRG) and activation of spinal microglia in the neuropathic pain. In other models of neuropathic pain, increases in reactive oxygen species (ROS) and decreases in GABAergic tone in the dorsal horn have been described. A disturbed balance between inhibitory and facilitatory impulses in the dorsal horn is the main mechanisms for pain hypersensitivity. The studies proposed will test the hypothesis that: pain in HIV/HAART neuropathy results from the production of proinflammatory cytokines that result in the release of ROS leading to a reduction in GABAergic tone in the dorsal horn of spinal cord. We propose a series of experiments with models of HIV/HAART neuropathy to: 1) define the changes of TNFa, ROS, glutamic acid decarboxylase (GAD) and GABA in DRG and spinal cord in HIV-associated pain; 2) critically test the relationship between elevations in TNFa and ROS; 3) examine the relationship between ROS and GABAergic tone; and 4) determine whether HIV-associated pain can be reduced by herpes simplex virus vector-mediated gene transfer to increase anti-inflammatory cytokine and inhibitory amino acid to DRG and their neurochemical mechanisms. The results of these studies will provide important insights into the pathogenesis of neuropathic pain in patients with HIV/AIDS, and may provide preclinical evidence for a novel therapy.